5The Tripartite Intelligence Alerts Agreement

In December 1954, NATO adopted a new strategy that relied heavily on nuclear weapons. The new strategy raised thorny questions about who would make the decision to launch nuclear weapons, and the British, with the Canadians, scrambled to find a way into the American nuclear decision-making process. The main effort lay in coordinating national and international indications intelligence programs, which led directly to the Tripartite Intelligence Alerts Agreement of 1957.

In early 1954, the NATO allies embarked on a study to support the development of a new strategy “in light of the effect of new weapons.” The resulting strategy document adopted at the December NATO Ministerial meeting, MC 48, was based on the assumption that “Soviet aggression will take the form of a surprise atomic attack aimed at the sudden destruction of NATO’s atomic capability.”¹ To prepare for this possibility, MC 48 directed NATO authorities to “plan and make preparations on the assumption that atomic and thermo-nuclear weapons will be used by the NATO forces from the outset” in any future major war.²

In the lead-up to the December Ministerial meeting, British foreign secretary Anthony Eden feared that the allies might consider “the difficult question of the authority for Saceur to use nuclear weapons.”³ There was, in fact, nothing in MC 48 that would allow NATO’s Supreme Allied Commander, Europe (SACEUR) to use such weapons on his own authority. Yet in 1954, NATO’s Standing Group had also written a paper detailing a proposed series of operational alerts which would delegate specific powers to commanders upon the declaration of different stages. Eden and his officials were concerned by the draft alert system (which they seem to have mistakenly believed had been approved), which made “it theoretically possible for NATO Commanders to begin a thermo-nuclear defensive war on their own authority.” According to the British reasoning, NATO might adopt, via MC 48 and the NATO alerts system, a system that would create “great political difficulty.” If Ministers were asked in the House, they would have “to admit that circumstances existed under which atomic warfare could be launched without governmental sanction.”⁴ The British, unwilling to “abdicate their responsibility on so grave an issue as this,”⁵ requested a meeting on the periphery of the Ministerial meeting with Secretary of State John Foster Dulles and invited the Canadian secretary of state for External Affairs, Lester B. Pearson, to join.⁶

On December 16, Dulles, Eden, and Pearson, along with advisers, met to discuss Eden’s concerns.⁷ During the meeting, in discussion over just how SACEUR would come to be authorized to employ atomic weapons, Dulles said that the “three or four governments who would carry the main load in war”⁸ — he obviously intended to include Canada among these — might try to find a formula; and he hoped that discussions would be held outside the council.⁹ None of the records of the meeting identify any specific agreement to coordinate an intelligence indication system, and none mention “intelligence alerts,” as the phrase was not yet in use in 1954. Nonetheless, this was the crucial conversation that would begin the diplomacy that led to the implementation of the Tripartite Intelligence Alerts Agreement. Based on later references to the meeting, it is clear that Dulles was reiterating earlier US pledges to Canada and the United Kingdom to consult before use of atomic weapons.¹⁰ Following the logic of US strategy — that those weapons would only be used if a Soviet attack seemed imminent — there seems to have been some suggestion of the three establishing a system to communicate any indication of impending hostilities.

A few days after the meeting, Eden followed up with Pearson, alone. Eden wished to once again “express his anxiety over the steps which should now be taken to work out ‘alert’ procedures by which action could be coordinated in an emergency.” (Eden used “alert” in the sense of an operational alert.) He told Pearson that he was going to ask Sir Norman Brook, the Cabinet secretary, to think about the problem and then discuss with Washington and Ottawa “in the hope that the three governments could agree on plans.” Pearson told Eden that the Canadians had already “worked out some technical arrangements with the United States in regard to ‘alerts’ and emergency action” for continental defence, and that this might offer a way forward.¹¹

Injecting Words of Caution

Eden wasted little time following up. The JIC (London) met on December 23, 1954, and the chair, Dean, explained that, in light of the Paris discussions and MC 48, it was “necessary at the earliest possible stage to examine the machinery whereby an agreed U.S./U.K./Canadian evaluation of urgent indicator intelligence could be reached and passed to the highest political levels in all three countries.”¹² After the holidays, on January 4, a “high level meeting in the Cabinet Office” led to the UK JIC being tasked with studying a “system of evaluating urgent indicator intelligence.”¹³ Dean cabled G. G. “Bill” Crean, the chair of Canada’s Joint Intelligence Committee, to tell him Eden had put in motion a program to draw up requirements for more “expeditious handling” of “indicator intelligence,” and “linking up what you might call our indicator centre with corresponding organisations in Ottawa and Washington.”¹⁴

Before the British could consider linking up centres, however, they needed to carefully consider their own indications and watch system. The British system, built around the pre-existing schedule of weekly meetings, was very similar to the early 1950s set-up in Canada. In standing procedure, any and all “significant intelligence” items were reviewed on Tuesdays by the heads of sections meeting in their regular place, the Joint Intelligence Map Room, on the fourth floor of the Ministry of Defence. The product of this effort, the “Weekly Review of Current Intelligence,” was then considered by directors of intelligence at the Thursday meeting of the Joint Intelligence Committee before being passed to the Chiefs of Staff and the ministers.¹⁵ There were provisions for urgent intelligence received both during and after working hours, but the British system was “not at present designed to meet a situation in which a surprise attack develops in a matter of hours.”¹⁶

At a JIC meeting in early January meant to fully examine the current practice, Ralph Murray of the Foreign Office questioned whether the existing machinery “was capable of operating fast enough under the circumstances envisaged by the latest thoughts on global strategy.” In the worst case, he noted, warning of an attack might precede an attack itself by only thirty minutes. London might likely get more time; several hours of warning, perhaps, and in more favourable circumstances, “a period of mounting political tension extending over days or even weeks.” A brigadier from the War Office agreed that the current system was unlikely to operate “really rapidly in an emergency.”¹⁷ After business hours, the War Office staffed a Duty Officer but not an intelligence officer who could respond and evaluate a warning. The RAF, on the other hand, had an intelligence officer on duty all day and night, but the RAF system was meant to deal with “an attack of which the first warning would be the appearance of hostile aircraft on UK radar screens.”¹⁸

Indeed, a Soviet attack in which complete surprise was achieved would “probably first be noticed on an Allied radar screen.” In such a situation, with an attack in the offing, the British would consider this to be an “operational matter” and so there would be no need for an intelligence alert.¹⁹ In a case where warning of attack was received with less than an hour, the “question of obtaining a Ministerial decision as a result of such a warning was, anyway, entirely academic.”²⁰

The British could imagine other scenarios, however, where there might be an advanced “indicator” (which they defined, using the same root, as “a measure which may be a significant indication of Soviet precautions for war”). There were three possible categories here: Soviet preparations meant to bring “operational units and facilities to immediate readiness for war,” those indicating that “the whole nation [was] being prepared for war in the very near future,” or those “indicating long term preparation for war.”²¹

The JIC (London) agreed that they required “accelerated evaluation machinery” which could enable Ministers to make decisions “within a few hours” in case indicators appeared. The machinery could be initiated, perhaps, by the “issue of a codeword.” They decided it should be possible for directors of intelligence at their Thursday JIC meeting “or in between meetings if necessary” to be authorized to call an “intelligence alert” by which “pre-planned arrangements should come into force” — i.e., the “Heads of Russian Sections,” officials responsible for the Soviet Union — would remain in office day or night, ready to evaluate and analyze.²² Ultimately, the British decided to avoid a codeword system as the group was so small, but they did implement a series of procedures by which special meetings could be called, offices staffed, and notice passed on to ministers.²³ (They would suggest a codeword for the tripartite system developed years later.)

It was for such a time that warning was available — not Soviet bombers on radar screens, but indications of preparation for war — that the British procedure was meant to evaluate. This desire seemed reasonable at face value. But discussions in the JIC make clear just why the British believed they needed to understand and evaluate early warning. These warnings would mean nuclear war.

In discussion, the JIC agreed that one of the most important decisions was when such “warning of attack” would be passed on — “expressed laterally” — to officials outside of the “intelligence machine.” The crucial question was whether it could be “emphatically ensured” that passing on an “intelligence alert” did not lead to “precipitate operational action” before ministers had so authorized.²⁴ Unlike operational alerts, the “calling of an ‘intelligence alert’” was to be a “purely a precautionary measure.”²⁵ It was in these conversations in London that the notion of an “intelligence alert,” separate from the operational alerts previously discussed, was coined.

British concern with indications intelligence and the need to differentiate intelligence alerts from operational alerts was directly connected to co-operation and interaction with the United States. The British knew that the Americans had built their own “Indications Center,” and it was precisely this that was so important to the British. The British needed their own system and machinery for evaluating indications intelligence if they “wished the Americans to allow [them] to maintain a working liaison with their Indications Centre.” And the British desired this very much, specifically “in order that we might inject words of caution into its [Washington’s] counsels.”²⁶

The Canadians, for their part, had sought to build relationships and reach agreements with the Americans meant to constrain any nuclear impulse in Washington until the Canadians could weigh in and agree on US decisions.²⁷ As Crean, chair of the Canadian JIC, explained to Dean, chair of the UK JIC, what the Canadians were “really concerned about is whether we shall receive expeditiously ‘hot’ items of intelligence, which might lead to an ‘alert,’ before we actually receive the ‘alert.’”²⁸ The Canadians did not want to be told by the president, let alone told by the US Air Defence Command that an “alert” was necessary, “without slightest knowledge of the information upon which their decision has been based.” Because an operational alert was so tightly coupled with the authority for nuclear war, the Canadians needed to find a role for themselves in the process leading to an alert. Crean, of course, as chair of the Canadian JIC, had his own parochial interests, too. For the JIC to have any real meaning and use in a crisis, it should “have the opportunity of making an assessment of all information which might lead to an ‘alert’ before the ‘alert’ actually takes place.” This was an important issue for JIC on its own, but also a larger problem, “one of principle involving the Government.”²⁹

For the Canadians, it was clear that the US Air Defence Command had been calling operational alerts (which, because of coordinated air defence systems resulted in alerts at Air Defence Command in St. Hubert, Quebec) “without telling us precisely the information which has led them to take such action.”³⁰ This was a problem strictly in the sense of continental defence. The adoption of MC 48 and NATO’s plans for an alert system magnified the issue for the Canadians, just as it had for Eden and the British. As Crean wrote in January 1955, it was not the “alert” system itself that was of greatest consequence. No matter how good the system was for implementing operational alerts, the “procedure depends fundamentally on an assessment of the information which leads to an ‘alert.’”³¹ It was now “more than ever important” that JIC (Ottawa) have from both London and Washington “by the most rapid means possible all information of a kind which might lead to an ‘alert’ so that an intelligence assessment of the information may be made here.” Without any such arrangements, “we shall, of course, be at the mercy of the operational commands” in the UK and the US.³² NATO, similarly, would be “at the mercy of . . . the United States and United Kingdom.”³³ If, however, Britain, the United States, and Canada could reach an agreement “tripartitely,” it would “go a long way to ensure that NATO commands receive properly evaluated intelligence from national staffs.”³⁴ Crean’s point, implicit but important, was that Canada should not leave these decisions up to others, even its closest allies.

At the 1955 Conference of Commonwealth Prime Ministers in late January and early February 1955, Eden passed on to Pearson the results of the UK Working Party that Sir Norman Brook had been chairing, and in particular a paper on “Possible Stages of Action when Indications of Major Russian Aggression are Received in Good Time.” The paper set out “the stages which ought to be completed if time allowed” if there were evidence of impending Soviet aggression specifically against the NATO area.³⁵

In the first stage described by the British, a proposed “London Indicator Centre” (what would be, the paper assumed, an adapted version of the current Joint Intelligence Committee organization) would receive a piece of information that might mean the USSR was making preparations for war. The London Indicator Centre would then contact its counterparts in Ottawa and Washington, ensuring “‘indicator’ experts of the United Kingdom, United States and Canada are fully in touch with each other on the matter.”³⁶ To be effective, new physical communications systems, especially across the Atlantic, would need to be used. In addition, to discuss the information between the three nations’ indications centres, relevant ministers, secretaries, and prime ministers and president would be consulted. If the three governments came to the “conclusion that war probably cannot be averted,” the British would approach Commonwealth governments while the British and Americans, with Canadian agreement, would approach the French — thus allowing the three Standing Group nations to approach the rest of NATO. The British paper noted that if the French “fail to make up their minds” the British and Americans would go straight to NATO without the French. The result of any approach to NATO would be a meeting of the North Atlantic Council and the approval of NATO’s operational “alert” measures.³⁷

The Canadian JIC not only found the British ideas acceptable, but advised the Chiefs of Staff Committee that “unless Canada takes part in some arrangement of the kind suggested, the Canadian Government may not receive the intelligence information on which to base conclusions regarding NATO alerts, answers to urgent requests concerned with Strategic Air Command action and with continental defence or other arrangements, or decisions required in other situations.”³⁸ The chair of the Chiefs of Staff Committee took the issue to the minister of National Defence, who agreed that “an indicator system as suggested in the United Kingdom paper should be set up in Ottawa for the exchange and evaluation of ‘Hot’ information with both the United Kingdom and the United States.”³⁹

The Canadians had some quibbles with details. The British paper was limited to major Soviet aggression against the NATO area. The Canadians thought there was a good case to include “the whole question of aggressive action by all possible enemies.” This was partially a function of the fact that Canada would serve as a base for SAC operations against the Soviet Union even in a war begun elsewhere, say in Asia. The Canadians also warned the British to scrap the reference to Commonwealth countries.⁴⁰

Ottawa relayed its interest in the British paper and suggestions for its improvement back to Dean, in London. The Canadians in London were instructed to make explicit that “the consequences of failing to take part in such a procedure might leave us [Canada] in the position that the Government might have to take a decision without full knowledge upon which such a decision should be based,” and that Canada would “be left in a worse position on the exchange of intelligence than we are at the present time.”⁴¹ Crean noted the Canadian interest in adapting the procedure to a global scope, and in particular Canada’s interest in “possible communist aggression anywhere, including possible Chinese communist aggression.”⁴² Were the British limiting their scope for tactical reasons, given US sensitivity to discussing Asia?

Dean explained that, indeed, the focus on NATO was tactical: that the British would start with NATO and then consider adapting the system to cover the Middle East and Asia. He thought that if the paper was “presented to the Americans as something springing directly, as it in fact did, from the last Ministerial meeting in Paris, it should be possible to restrict discussion to the NATO angle.”⁴³ What was more, Dean said, the NATO context was actually the most complicated of all global scenarios, given the role and authority (however unclear) of the SACEUR. If the three could find a solution for the NATO context, surely they could solve any other problem later. The Canadians and British were concerned about how they would staff any such organization to implement the procedure, as both ran rather threadbare intelligence establishments relative to the US. They hoped to solve that problem later (while assuring each other that not too many extra officers would be required to set up such a system). Dean agreed to scrap the reference to the Commonwealth countries while not, “of course, regard this as limiting our right to consult and inform Commonwealth Governments.”⁴⁴

Dean next considered how to approach Washington. He worried that a joint approach to the Americans would imply that London and Ottawa had come to conclusions without the Americans. Instead, he proposed that the British and Canadian ambassadors make separate approaches to John Foster Dulles. The British ambassador would present the corrected paper seen by the Canadians, and both the ambassadors would suggest the time had come to follow up on the discussions in Paris.⁴⁵

At this stage, Dean assumed the result of these approaches would be a political-military discussion involving British Ambassador Roger Makins, head of the British Joint Staff Mission (BJSM) in Washington General John Whitely, and officials from the CIA and the Pentagon. Makins could present the British paper as a basis for discussion and also mention that the Canadians and British had taken advantage of the Commonwealth Conference to discuss intelligence matters.⁴⁶

The Canadians ultimately agreed with Dean’s limitation to the NATO area, allowing that the first goal should be a “practical and speedy procedure preliminary to and in support of the NATO alert system” upon which could later be built “a parallel procedure for other areas of the world.”⁴⁷ They agreed that the British Ambassador would go first, and that Arnold Heeney, the Canadian ambassador, would follow up (he did so on April 29, 1955).⁴⁸ Optimistically, Dean hoped to have discussions before the next North Atlantic Council ministerial meeting.⁴⁹ In fact, discussions would continue for years.

A Canadian Indications Room

Throughout the first half of 1955, the US Watch Committee had been “materially stepped up as a result of a constantly increasing danger of devastating damage to the U.S. in the event major aggression caught us flat-footed.”⁵⁰ Nonetheless, there was still pessimism about the ability of the US to gain warning of a Soviet attack. Estimates of growing Soviet air capabilities made the National Security Council “somewhat more pessimistic than we were last year regarding our ability to give advance intelligence warning of surprise air attack.”⁵¹ If the Soviets were to attempt a major surprise attack from forward bases, it might be detected with a general warning of several days and a specific warning of a day or less, eighteen to twenty-four hours. A smaller attack, if carefully planned, could be launched in 1955 “with no assurance of specific advance warning to U.S. Intelligence (apart from that provided by early warning radar).”⁵² The race for indications continued: Soviet security and counter-intelligence measures meant there was no improvement in human intelligence collection capabilities within the Soviet Bloc — a place next to impossible to run spies — but the US had seen “considerable improvement in the collection of intelligence data through technological means . . . together with increasing use of aerial reconnaissance.”⁵³

Also in 1955, JIC members in Ottawa remained ambiguous about the value of indications intelligence. In February, the DNI, for instance, remained skeptical of paying “too close attention to ‘indications’ intelligence factors which can be fed into a machine to produce a mechanical answer” which “may well be misleading.” An “indications of war display” would only be effective if “the enemy is planning a deliberate war for an established date, when the progressive build up in armaments and in political tension could be watched and a danger point agreed on.” What Canada needed in peacetime, he thought (somewhat contradictorily), was “a well-equipped map-room with an adequate supply of relevant and up-to-date charts and diagrams.”⁵⁴

The DDNI, for his part, told the DNI that he did not think it was the right time to decide on an indications room for Canada. He knew the JIC (London) was studying the issue but was “undecided as to whether a list of indications is of much value” except in a gradual build up like the DNI had noted. Now, however, the Americans had an indications room “in actual operation” and since information in the Canadian room would “largely derive from UK and US sources,” Canada should postpone discussion until “the US room can be seen in operation and the UK make up their minds.”⁵⁵

The US room was, of course, up and running and it was impressive to outside observers. Just as the British were reviewing their set-up, the Canadians were trying to learn more about the US National Indications Center that had been in operation since January 24, 1955.⁵⁶ The NIC was essentially the staff of the Watch Committee. While the USAF had kept their “Indications Board,” there was no other indications staff in Washington, DC — the NIC was it.⁵⁷

The Canadian JIB Liaison officer in Washington visited the NIC and was briefed by its Director, J. J. Hitchcock.⁵⁸ Unlike in Ottawa and London, space did not seem to be an issue for the NIC. It occupied a suite in the basement of the Pentagon, which included a conference and briefing room for the Watch Committee, a reception room, and offices for the director and clerical staff, including cubicles for representatives from all the various US intelligence agencies assigned to the NIC. The salaries of representatives were paid for by their respective service, and “Rations and Quarters” were provided by the Air Force, even though the NIC itself came under control of the IAC.

The NIC itself received information from all agencies as well as ticker tape updates from the Associated Press and the United States government’s Foreign Broadcast Information Service (FBIS). Certain incoming cables from collectors in the field were also automatically copied to the NIC. The service officers were charged with maintaining links to their home agencies so they could receive assistance in evaluating incoming material. The links also ensured they received any raw intelligence that should have been sent to them in the NIC but was not.

The officers worked twenty-four hours a day to evaluate incoming material, and the NIC could call an unscheduled Watch Committee meeting in a “tense” situation. The Watch Committee then would agree “on a joint interpretation of events to be disseminated to the proper authorities.”⁵⁹ During the First Taiwan Straits Crisis of February 1955, in which the US Navy evacuated Chinese nationalists from the Tachen islands, the Watch Committee had been called and agreed to a joint appreciation in two hours.

The NIC, ostensibly, was to be responsible for monitoring indications intelligence from around the world to study long-term indications of war. But in its early stages, Hitchcock explained, it was impossible to begin a wholesale indications program, and so the NIC was focusing on two crisis points: conflict between the People’s Republic of China and the Republic of China over islands in the Taiwan Strait; and Berlin. This would help the NIC establish an efficient system, and starting so small was safe, Hitchcock said, as “it is a general appreciation that a major war is unlikely to occur in the near future.”⁶⁰

The JIB officer noted that the Watch Committee’s room in the suite was “elaborately equipped with visual aids.” Hitchcock assured him, however, that he was more concerned about collecting, evaluating, and disseminating information than “developing any mechanical mathematical or other alleged foolproof system for measuring the temperature of international tension.” Hitchcock’s major problem, he said, was getting the co-operating agencies to pass him only the “right kind” of material, rather than “swamping him with irrelevant documents.” He hoped that field collection would improve, because attachés and others were not currently looking for indications intelligence — or at least not the right kind, as he saw it.⁶¹

Hitchcock himself disparaged the regular Watch Committee report, which he thought simply a survey and “not a barometer of the imminence of war or an indications report at all.”⁶² He hoped to convince the Watch Committee to include an indications annex that might one day overtake the survey itself. This was news to the Canadians who received the Watch Report. It was obvious that the exchange of regular intelligence products would not be the best source for providing indications of war. These things would be detected and then made manifest in an emergency or “crash” meeting of the Watch Committee — and Canada did not receive reports of these meetings. Just as Crean had written to Dean about “crash” meetings of the JIC (London), it was obvious that what really mattered for understanding US indications detection was information about “crash” meetings in the Pentagon. Hitchcock could not provide such information on his own authority. He suggested that Canada could only receive such warning if Allen Dulles and Crean could reach a new agreement, one that would provide the NIC with instruction to release material not currently sent to Ottawa.⁶³

In March, the chair of the Chiefs of Staff asked the JIC to describe just how a Canadian system for handling indications intelligence would work.⁶⁴ Less than a week later, the JIC had approved a new paper, JIC 135/1(55), “Indications Intelligence,” advising of the JIC’s conviction that an indications room should be established in Ottawa.⁶⁵ The paper sketched out how such a system would operate: the new Canadian Indications Room would be responsible for passing intelligence received from allies to the directorates of Intelligence for evaluation, and, if necessary, call a “crash” meeting of the JIC. There was no twenty-four-hour watch suggested, as the JIC assumed none would be necessary unless there existed an obvious escalation of tension. Nonetheless, an indications room was necessary, they argued, as “Canada must develop its opinions as to the significance”⁶⁶ of items of intelligence that the US and UK believed worthy of action.

The JIC expected that Canada could avoid having to “duplicate the scanning of the bulk of United States and United Kingdom source material” and focus on assessment or analysis. While both the US and UK were sensitive about “third nation” information — that is, information from another state — the Canadians assumed that in any new system, they could expect to see third-nation information if Washington or London thought the intelligence “to be of indications significance.”⁶⁷

The system, as envisioned in JIC 135/1(55), would require consideration of only a “small volume” of information. Contradictorily, and cutting against the Canadian assumptions that the primary work would be done by the other allies, the JIC noted the “increasing probability that an indication will turn up in Canadian source material.”⁶⁸ This was likely an allusion to Canada’s responsibility for collecting signals intelligence in the Soviet Arctic.⁶⁹ This all added up to mean Canada would need its own system for detecting and referring indications intelligence it collected.

The JIC expected that the current “Hydra” system would solve the problem of communication. (For more on the Hydra communications hub, see Chapter 6.) The system had been in place for years, had agreed reservations on its use for specific users, and was “designed primarily to carry intelligence traffic.” If the Hydra link in the UK at Government Communications Headquarters (GCHQ) was extended to the Ministry of Defence, and either the Hydra link at the National Security Agency (NSA) or the Canadian link at the Ottawa Wireless Station (located in Leitrim, just south of Ottawa) could be extended by line to the Pentagon, all three indications centres could be linked directly.⁷⁰

Waiting for an American Response

The year ticked away with no response from the Americans to the British and Canadian overtures made in April 1955. In early July, UK foreign secretary Harold Macmillan and Patrick Dean visited New York and met with John Foster Dulles. They asked American officials about the alerts issue. While Livingston Merchant of the State Department was “hopeful” the US might come to an agreement, a CIA official let on that the CIA “was not very favourably disposed.”⁷¹ A few weeks after the New York conversations, Makins checked with the State Department, who explained the delay had been caused by the slow response of the US permanent representative to NATO, George Perkins, who had been asked to comment on the British proposal.⁷² Canadian intelligence officials like Crean continued to hear nothing about alerts when they co-operated with the American colleagues on other matters.⁷³

Philip Uren, the Canadian liaison officer in Washington, asked Robert Amory, the deputy director of the Central Intelligence Agency, where things lay, but Amory knew nothing about even the initial “high level approach” by the ambassadors. After hearing from Uren about the proposed plan, Amory said he expected the answer would be “unfavourable,” partially because of a Pentagon-CIA spat. Amory also criticized the way the British and Canadians had gone about achieving their goal — that is, by raising it on a government-to-government level. Amory blamed the “British tendency to carry matters to the highest level unnecessarily.” Had Crean, as head of JIC, simply proposed the solution to Allen Dulles of the CIA, Amory said, “an effective liaison system with the NIC would now be in operation.”⁷⁴

As Amory came to understand the Canadian concern — that Ottawa might be left out of the analysis and decision-making process in Washington — he told Uren that the Watch Committee would never hold an emergency meeting on “a subject important to us [Canada]” without Amory personally informing Uren. And if the situation was “sufficiently serious,” Amory said, he was “personally prepared to over-ride rules concerning Officials” and, by implication, tell Uren what Canada needed to know. If the Canadian JIC held a similar meeting, Amory said, he hoped he would be informed in kind, “as the Canadian action would effect [sic] his attitude to the problem in Washington.”⁷⁵ The conversation is instructive, for it suggested the close relationship between Canadian and American intelligence officials. But it clearly relied on trust at the personal level. What could the Canadians expect with a different cast of characters? If Amory were absent? In a real crisis, would any American think of the Canada? The Canadians were seeking a system and a habit, rather than personal guarantees.

By October 1955, neither the British nor the Canadians had heard back from their approach in April. C. Burke Elbrick, the deputy assistant secretary of state for European Affairs, said that the matter was stuck up with Defense. All that the Canadians heard were “rumours and they are generally to the effect that the Pentagon does not like the tripartite scheme” probably because of the “usual conservative attitude of some military officers in relation to what they regard as sensitive intelligence.”⁷⁶ While this may explain part of the delay, it seems that neither the Canadians nor the British understood the full backstory of the formation of the Watch Committee (and thus the NIC), let alone the theoretical powers of the NIC to be informed of US operations around the world.

The Canadians, like the British, had expected the discussion to move quickly after their first approach to the Americans in the spring of 1955. While Ottawa had an “intrinsic interest in obtaining tripartite agreement in alerts procedures,” the Canadians also wanted to use discussions “as a point of departure for further talks on the broader questions of the circumstances under which nuclear weapons might be employed, tactically and strategically, not only in the NATO area but in any other part of the world.”⁷⁷ In particular, they had thought the tripartite discussion would connect with separate efforts to seek an agreement with the US on a system of bilateral alerts regarding North American defence. They had also not given up their concerns about precipitate American action in Asia, which would almost certainly end up involving Canada given the integrated nature of North America’s continental defences.

In late September 1955, in planning for regular high-level US-Canadian consultations, the Canadians considered whether to try and bundle these matters for discussion with the Americans, or to approach them individually.⁷⁸ Jules Léger, the USSEA, thought it “important to get on with” US-Canadian discussions regarding operational alerts in North America, “the exchange of indications intelligence on the Far East.”⁷⁹ But he also did not want to “prejudice” the stalled tripartite system that the British had proposed, with Canadian support.

As the Canadians were deciding how to move, Admiral Arthur W. Radford, the chair of the US Joint Chiefs of Staff, told the Canadians that the Pentagon had already “turned down” the proposal for a tripartite agreement but not yet told the State Department. Radford said that the Joint Chiefs of Staff were reluctant to agree to any form of consultation with foreign powers — the word “consultation” is important. It was already hard enough, Radford said, to obtain “co-ordination inside the United States Governmental machine, in particular between the Pentagon and the Central Intelligence Agency.”⁸⁰ Radford’s comments raised a host of concerns, but one effect was for the DEA to suggest that the Canadians begin discussing a bilateral alert system with the US and seeing how the conversation developed from there.

As the Canadians planned to move forward on bilateral issues, the US replied to the British paper on tripartite indications. On November 18, 1955, British and Canadian embassy officials were called to the State Department, separately, and given an aide-mémoire.⁸¹ This was a curious development because the Canadians knew that opinion within the US government was split. The Air Force and the CIA supported tripartite arrangements. However, the Army, Navy, and Joint Chiefs of Staff were opposed, and rumour had it the IAC had given a “unanimous negative answer to the proposal.” Crean assumed that “somebody higher up the line modified the position taken by the IAC.”⁸² Amory of the CIA let on that he thought the IAC had not understood what the British and Canadians were asking for: IAC members had variously interpreted the British memorandum as a request for more intelligence sharing and consultation.

No matter the process that led to the aide-mémoire, Glazebrook (now at the Canadian embassy in Washington) observed that it was “much more encouraging than all the gossip indicated it would be.” Despite the document’s lack of clarity, it “seems to add up,” Glazebrook thought, “to some kind of acceptance of the original United Kingdom proposal endorsed by us.”⁸³ The American reply also suggested that discussions could move forward on a tripartite basis between the State Department and the British and Canadian embassies in Washington.⁸⁴

The American aide-mémoire was a “rather confusing” document as read in London and Ottawa, seemingly hastily assembled. While it noted American concerns and errors in the British document, it also seemed to suffer from sloppy drafting.⁸⁵ The Canadian JIC considered the aide-mémoire paragraph by paragraph, and were put off by the very first paragraph that seemed to declare an American authority to defend Canada without consulting the Government of Canada.⁸⁶

The greatest concern to the Canadians was that the US aide-mémoire suggested that the existing system for the exchange of intelligence was suitable. This was a sticking point for the Canadians. They were not convinced that the ad hoc arrangements that had been established for sharing intelligence across the border were “clearly established or expeditious.”⁸⁷ This was important to the Canadians for both bilateral and trilateral procedures for intelligence warning. Canada had no “direct liaison” with either the Watch Committee or its Indications Center.”⁸⁸ The aide-mémoire seemed to suggest the Americans thought there were effective procedures for getting in touch with Canadians in a crisis. However, what the Canadians really sought was to be able to see “the essential intelligence”⁸⁹ that the Americans were considering. What the Canadians wanted was a system of “direct contact” — including a direct line of communication — between Canadian and US intelligence (be it the NIC, Watch Committee, or IAC). The establishment of such a system, they expected, “should produce the assurance in a time of crisis.”⁹⁰

The Americans had clearly misunderstood parts of the British paper while the Canadians and British were less than enthusiastic about the US aide-mémoire. Although the Americans agreed to talks, just how to move forward was unclear. A simultaneous US-Canadian meeting and subsequent follow-up on continental defence issues would provide the solution.

Bilaterals and Trilaterals

Around the time the Canadians received the American aide-mémoire, they were considering how to approach the discussion with the US regarding bilateral alerts in support of North American continental defence. A Canada-US agreement signed in 1952 that allowed ministers to approve US overflights of Canadian territory was of “limited usefulness given interdependence of air defence.” What was needed now, wrote Léger, was “an arrangement for the exchange and evaluation of strategic information of a kind which might lead to a decision to take emergency measures or even to go to war.” Any such agreement would have to be accompanied by “a firm understanding of the necessity for consultation at the highest political levels of the two governments on the action to be taken as a result of that information.”⁹¹

Ultimately, this boiled down to the need for the Government of Canada to ensure it had “the information it would need to arrive at independent conclusions in an emergency regarding the operation of the continental air defence system and the deployment into or over Canada of the Strategic Air Command.”⁹² This was a matter of continental defence, but the notion of continental defence and general war were largely indivisible.

Now that the Americans had responded to the British paper with their aide-mémoire, Léger thought it a “propitious” moment to raise the matter of alerts on a bilateral basis with the United States. The Canadians would keep the British in the loop by telling them they were going to discuss alerts with the Americans. But the Canadians would press for “a firm understanding”⁹³ on procedures in the air defence field and separate from the connection to the NATO system of alerts on which the British paper had been based.

As a starting point for the discussion with the US on bilateral alerts, the DEA drafted a formula of four points by which this process might work, and they shared it with the Americans. According to the proposed formula, if the US or Canadian intelligence authorities concluded that “there was a likelihood of hostilities occurring in which North America would likely be attacked,” the Governments would “automatically pass to one another all relevant information,” including background information and assessment. This would ensure that if consultations “at a higher level” were required, “Ministers would be fully in possession of the necessary facts.” ⁹⁴ This formula, the Canadians assumed, “might usefully be applied, and probably later, to the tripartite scheme.”⁹⁵

The formula, however, did not dispel the cloud of confusion in Washington. When Glazebrook tried out the bilateral formula on State Department officials on January 24, he found them on the defensive. The Americans — C. Burke Elbrick, Outerbridge Horsey, Park Armstrong, and H. E. Furnas — began pushing back by outlining the present exchange of “alert type” or “indicator” intelligence between the United States and Canada: The United States passed its Watch Committee reports to the Canadians immediately upon printing. The RCAF officers at Colorado Springs had access to intelligence, and a special telephone line linked the Joint Air Defence at Colorado Springs with Canadian Air Defence Command. The radar screens that would provide alert intelligence in the north were jointly operated, and the NSA and CBNRC exchanged intelligence. Where, the Americans asked, were the gaps?

Glazebrook sensed that the Americans believed the Canadians were asking “for a more complete exchange of intelligence.”⁹⁶ He denied this immediately and recognized the misunderstanding — not only in the State Department, but no doubt this was the source of confusion in IAC, too. He explained that, “leaving aside the normal exchange of intelligence, there did not exist, in our opinion, a defined arrangement by which the United States and Canadian intelligence authorities could quickly correspond on an agreed channel and, where necessary, exchange views.”⁹⁷

Armstrong replied that if there was enough time for the Watch Committee to meet and appraise the situation, the results would be passed to Ottawa “by the quickest means.” But this was the sticking point. Glazebrook noted that “the ‘quickest means’ were at the moment uncertain and not necessarily quick.” The Americans just assumed they would pass the information to the “Canadian liaison officers.”⁹⁸ By this, they must have meant Uren — but he was not on duty 24/7, and he also had no ready channel of communication to Ottawa.⁹⁹ Furthermore, as the Americans acknowledged, some types of intelligence could not be transmitted over regular channels. The State Department men seemed to recognize that there was not an effective system in place, and Glazebrook believed Armstrong would now champion the issue within the IAC. Armstrong promised to speak with Director of Central Intelligence Allen Dulles, and the Canadians would wait until State had spoken with the CIA. It seemed to Glazebrook that the State officials were seeking to leave the American military out of the process, since bringing in service officers “would mean endless comings and goings with the Pentagon.”¹⁰⁰ Ultimately, the meeting served to identify “the bogey man lying behind the United States Aide Memoire on the tripartite scheme.” If the Canadians could make more headway on the bilateral plan, “there may be some hope of sorting out the tripartite scheme later on.”¹⁰¹

A month later, Crean travelled to Washington where he had several meetings each with Amory of the CIA and Park Armstrong of the State Department. Crean continued to make the case for a bilateral system between Canada and the US, pressing for a direct line between the NIC and the Canadians Indications Room. Ottawa, he told his interlocutors, was “anxious to eliminate from the system any built-in delays such as getting cypher operators or officers out of bed before a piece of information could be passed in either direction.”¹⁰²

Amory, who did not seem to have discussed the issue with his State Department colleagues, was less sanguine than the officials Glazebrook had met with a month earlier. He warned that the IAC was unlikely to accept the formula as it stood. The president, he said, would “almost inevitably initial it,” indicating his initial agreement. But then the services will “see all sorts of objections and advise him differently.”¹⁰³ Amory again advised that the direct link between the NIC and Canadian Indications Room should be dealt with as a matter for Crean and Allen Dulles on their own, to be treated as a “nuts and bolts” problem between their two offices rather than a government-wide matter (which would bring in the military in Washington).¹⁰⁴

Armstrong, too, seemed to backtrack from the earlier meeting. He hypothesized about the difficult case where a “delicate item” of intelligence that might be “difficult to pass” along a communications channel that ended in another state. He urged the Canadians to increase the liaison staff, but Crean rejected this idea, not wanting to staff the Washington embassy with officers waiting around for an emergency. Crean explained that he knew there might be information from “double agents and that type of thing”¹⁰⁵ that the US could not pass, and “obviously, one would not put on the link what one freely wished to trade.”¹⁰⁶

The most crucial lesson Crean learned from his trip was Amory’s observation that the formula as it stood now seemed to combine the issue of “consultation between governments” on one hand, and on the other, “the problem of exchanging information during a period of crisis which might lead to a Declaration of an Alert by either government.”¹⁰⁷ Amory asked: Did the Canadians expect to receive the Strategic Air Command’s operational plans? (Amory’s question, likely unbeknownst to Crean, mirrored the earlier discussion in the US about the relationship between US operations and indications intelligence.) Amory suggested that if the Canadians could separate these two issues, there may be room to move forward. But “consultation” was the main point of contention. It would ultimately stand in the way of easy agreement on exchange of information, as the United States government was unlikely “to bind itself under any formula which required it to consult another government before taking action itself.”¹⁰⁸ This would have important implications for the tripartite approach, too.

By April 1956, the Canadians decided that the informal discussions with the Americans on the bilateral issues had essentially run their course. In March, the Canadian director of air intelligence met with the US Air Force director of intelligence Brigadier General John Samford. The American warned he was “not optimistic that early agreement would be reached to exchange indications.” The problem was that within the US Watch Committee itself there were significant problems, including “considerable reticence” by agencies to “table all indications available.” Several agencies preferred only to table their evaluation of indications, rather than the data itself. Facing “this difficulty at home,” Samford “could not see how a free exchange of indications with Canada was possible.”¹⁰⁹ The Canadians were not sure whether to take these comments at face value or as a hint that a request would be rejected.

Nonetheless, the Canadians decided they would make their formal approach on the bilateral issue.¹¹⁰ In the Cabinet Defence Committee, Pearson and the Minister of National Defence Ralph Campney argued that the “main purpose” of such an exchange was “to ensure adequate consultation before a recommendation was made in either country for the calling of an [operational] Alert.” Such an alert was synonymous with imminent war; in the event an alert was called, the United States would “evacuate the principal government activities from Washington.” There was, the Cabinet Defence Committee agreed, “very little that was more important than the actual calling of an Alert.” As a result, “everything should be done to ensure that this step was taken only if the necessity was clearly demonstrated.”¹¹¹ Ultimately, the Canadians were seeking to build a system to ensure that the Government of Canada was a) “in a position to decide whether an alert should be called” and b) to ensure that Ottawa was “consulted by the United States Government.”¹¹² While the ministers’ statements dealt exclusively with bilateral Canadian-American relations, the Canadian thinking paralleled the tripartite issue and mirrored the earlier British thinking about inserting wisdom into American counsels.

The Canadians took Amory’s advice to heart and drafted two formal letters, splitting the issues: one suggesting an agreement on the exchange of intelligence, and the other on consultation.¹¹³ The Canadians were prepared to argue that the United States should consult with them because Canada was willing to agree to joint operational control of air defence in North America.¹¹⁴ (Intriguingly, as will be seen, the Canadians did not make the same push for “consultation” when it came time to negotiate the tripartite agreement.) The letters progressed through the Cabinet Defence Committee, and Heeney presented the letters to John Foster Dulles in May 1956.¹¹⁵

The Canadians learned their letter on the exchange of intelligence had been approved relatively quickly, but discussion on the consultation issue dragged on inside the US governmental machine.¹¹⁶ While they were waiting for a response, Dean approached the Canadians about taking up trilateral issues again.¹¹⁷ The Canadians would have preferred if the bilateral problem were solved first, but ultimately decided it was better not to delay the tripartite issue because it was unclear when the US would respond to the bilateral letters. More importantly, as Glazebrook noted, the bilateral discussions both he and Crean had with US officials had “removed some serious misunderstandings which had clearly existed in the minds of United States officials”¹¹⁸ and should help move the tripartite discussion along.

In the summer of 1956, Dean visited the National Indications Center and discussed alerts in Washington. The NIC’s director explained to Dean that the “Americans had been rather frightened by our original approach [in spring 1955], because it had included suggestions for high-level policy discussions as well as exchanges of information through intelligence channels.”¹¹⁹ Upon his return to Washington, Dean, clearly influenced by both the Canadian experience and his visit to the NIC, suggested to Crean that the British and Canadians approach the Americans for tripartite discussion “on the exchange of intelligence only.”¹²⁰

It seems the matter was dropped by all governments in August 1956. The obvious reason is the Suez Crisis of 1956. Dean played a significant role in the US-UK minuet before the crisis. But in the aftermath of the Suez Crisis, the need for such a system became even more pressing for the British. Harold Macmillan became the British prime minister and US-UK relations experienced a significant rapprochement, including plans for the US basing of Thor missiles in the UK.¹²¹ Macmillan seems to have pushed hard for an agreement, and he used key meetings with Eisenhower in 1957 to force movement in the US bureaucracy. Crean knew that the British were willing to move forward on an intelligence-only basis but expected they may wish to return to the issue of “consultations” later. As the British were going to allow US bases in the UK, Crean expected “they might have a strong point and one which is in some measure comparable to our own position with respect to continental alerts.”¹²² The upshot of the Suez Crisis then, or at least the following rapprochement, was a stronger British interest in, and case for, a formal agreement to exchange indications intelligence.

In February 1957, Crean travelled to Washington for more meetings with Allen Dulles and other intelligence officials. In his meeting with Dulles, Crean asked about the tripartite issue and Dulles was unwilling to commit himself, saying only that he saw “no reason why the present time would be unsatisfactory” for resuming talks.¹²³

In a meeting with Charles P. Cabell, deputy director of the CIA, Crean suggested “an informal tripartite talk confined to civilians, for example, Dean, Allen Dulles, and himself.” Cabell, however, was resistant. He said that the British and Americans had their own agreement, as did the Canadian and Americans. There was “[n]o necessity to complete the triangle in a single document.”¹²⁴ Crean disagreed, pointing out that the US-Canadian agreement focused on continental defence only. Cabell responded that the United States did not interpret the agreement that way. Crean himself seems to have wondered whether it was all looked after under present arrangements, including the informal expansion offered by Americans. But, he asked, “are informal arrangements likely to work in a crisis?”¹²⁵ The trip to Washington only raised more questions.

Following his trip, Crean drafted a new formula to serve as the basis for a tripartite discussion. It was based on the US-Canadian bilateral intelligence-sharing formula, but applied to the NATO area.¹²⁶ If any of the three governments received information and concluded “there was a likelihood of hostilities occurring in the NATO area . . . it will keep the others informed.”¹²⁷ The intelligence authority that reached such a conclusion, be it IAC or one of the JICs, “will pass to one another automatically and by the most expeditious means all intelligence information”¹²⁸ used to reach the conclusion, as well as background information and assessments.

Crean passed his draft to the high commissioner in London, Norman Robertson, who in turned would pass it to Dean. Robertson’s comments on the draft formula are instructive, for they preview future Canadian concerns with the proposed system. The formula Crean sent was restricted (as had been the British working paper) to aggression “occurring in the NATO area.” Robertson thought this was problematic and wondered what would happen if there was information “suggesting that the Chinese Communists were preparing an attack on Formosa or even the off-shore islands.” Either scenario might result in a Soviet attack on North America “as a result of the operation of USA Defence arrangements with nationalist China on the one hand and the Soviet-Chinese Communist Alliance on the other.”¹²⁹ Furthermore, the British action at Suez, “without prior notification to the American or ourselves, illustrates the actuality of this problem. It is not inconceivable that such unilateral action by one or other party could if repeated lead to hostilities which might spread to the NATO area.” Ultimately, Robertson conceded, “perhaps there is no satisfactory cure for the problem” other than trying to consolidate “co-operation and confidence between gov[ernmen]ts.”¹³⁰

Dean received the new Canadian formula and also Crean’s report of his meeting with Cabell — including Crean’s assessment that the “CIA were very dubious about the desirability or necessity of a tripartite agreement.”¹³¹ Dean, who thought he had reached agreement with Dulles was “rather put out” by this “shift in American thinking.”¹³² Dean’s political masters, too, were “disturbed” and “very anxious to get on as fast as possible with establishing an agreed speedy three-cornered arrangement for exchanging intelligence.”¹³³ The ministers agreed that the issue should be raised, and officially entered on the agenda, when Macmillan and Selwyn Lloyd met with Eisenhower and John Foster Dulles at Bermuda in March 1957. As the British had begun their quest for a tripartite arrangement “largely on the suggestion of Foster Dulles himself,” they no doubt expected that if the issue was raised at the highest levels they could cut through “misunderstanding” of lower levels of the United States government.¹³⁴

Dean proposed that Crean and Allen Dulles meet immediately following the Bermuda conference.¹³⁵ The Canadians liked this idea because their earlier informal discussions in Washington had helped them prepare their formula and letters related to bilateral arrangements. Still, the Canadians agreed among themselves that this was “certainly not our proposal.”¹³⁶ They were ready to let the British take the lead, associating themselves with any meeting that came out of the British request.¹³⁷ For his part, General Foulkes was relieved that the British would not be sending an air intelligence officer, because otherwise he would have to send an officer and “no doubt that the Pentagon would insist on having three intelligence representatives.”¹³⁸

But Foulkes wanted to be sure the discussion kept to intelligence exchange rather than consultation. Radford had told Foulkes he was “not at all enthusiastic about United Kingdom – United States consultation on alerts.” Foulkes thought that any effort to expand the tripartite discussion to include consultation, rather than intelligence exchange, would negatively affect bilateral US-Canadian relations. The Pentagon, if “faced with trying to arrange a suitable mechanism for consultation on a tripartite basis they might raise so many difficulties that they may call off the arrangements for consultation on a bilateral basis.”¹³⁹

Given Cabell’s comments, Robert Bryce, the clerk of the Privy Council, thought the meeting might be useless. Still, he could not see “any objection to an informal discussion” between the three intelligence directors, if “only to discover precisely how effectively the two bilateral agreements may be expected to work.” He endeavoured to find Crean a seat on the government plane headed to Bermuda.¹⁴⁰ The British were successful in having tripartite alerts added to the agenda for Bermuda, but did not expect any decisions by Eisenhower or Dulles beyond confirming that representatives of the three states should meet again in Washington to discuss things further.¹⁴¹ Crean then found a seat on the British plane returning from Bermuda to Washington, DC, where the Canadian and British representatives would meet with Dulles.

The British, for their part, thought the Canadian efforts had “set a useful precedent for a tripartite procedure” and had taken on board the Canadian suggestions of focusing only on intelligence rather than consultation. The British committed to concentrating, “in the first place” on “agreed procedures for exchanging intelligence on Soviet intentions to attack the NATO area.” They would leave consultations — what they explicitly referred to as “the question of the political decision to authorize the NATO military commanders to use nuclear weapons” for later settlement.¹⁴²

Building a Canadian Indications Room, Part 1

During the tripartite diplomacy, the Joint Intelligence Committee had been giving attention to Canada’s ability to handle indications intelligence. In November 1955, the Americans had provided their response to the initial British and Canadian approach in an aide-mémoire. Glazebrook, in Washington, had heard rumblings about the Ottawa JIC’s plans to improve the Canadian Indications Centre. He thought the Americans would be watching these developments closely. Glazebrook knew that the Canadian indications effort was a “spare-time operation,” with rotating staff who had little training. If the Americans were to realize this, “there would be a good deal of puzzlement here [in Washington] as to why we had made a fuss about the thing at all.”¹⁴³ If the Canadians were to participate in a network of indications centres, they would need a properly functioning centre of their own in Ottawa.

Although the JIC had taken steps to try an indications project, the US aide-mémoire confirmed for the JIC that unless there existed “joint machinery . . . to exchange ‘hot’ information” with the Americans, the “whole basis of political consultation with the US in an emergency would be jeopardized.” This was a matter of “some urgency” and the Canadians agreed to put their “indications project” on a twenty-four-hour basis for a three-week trial basis beginning January 3, 1956.¹⁴⁴ Building on Amory’s advice to keep the connections between the US and Canadian indications centres on a “nuts and bolts” level, Crean wrote a letter to Dulles suggesting the “existing link between N.S.A. and C.B.N.R.C.” be used to connect indications centres.¹⁴⁵ Dulles then cabled A. J. Steele, the CIA representative at the US embassy in Ottawa, and agreed, saying that the terminal link for the US will be in the CIA, not in the NIC.¹⁴⁶ (This was to connect to Cabell’s office in the CIA, for Cabell was in charge of the Watch Committee.)

By May 1956, the JIC was convinced it was “essential to run a Joint Indications Room into which all Directorates and Departments represented on the J.I.C. will submit on a timely basis items which may lead to the opinion that war was imminent or in preparation.”¹⁴⁷ Throughout the month, the Chiefs of Staff Committee discussed terms of reference for a Joint Indications Room (JIR), considered the challenge of passing British information to the US and vice versa, and made plans to staff and support, administratively, a JIR that was established that month.¹⁴⁸ In July, JIC established a routine for “action in event of special message from Washington.” Including during off-duty hours, the routine essentially consisted of a lengthy list of who should call who to ensure a quick emergency meeting of the JIC.¹⁴⁹

By the end of 1956, the JIR was drafting a Weekly Indications Report, though its purpose was still debated. Bowen, from JIB, was pressing for “Indications” to be interpreted “in a broad sense to include developments which would have a significant bearing on the likelihood of war, e.g., to include changes of political leaning or military capabilities which would encourage or discourage parties to a dispute from the use of force.”¹⁵⁰ Bowen, who seems to have given the most thought to the theory and purpose, rather than just the technical arrangements, of indications intelligence, was “firmly convinced that by far the most important task confronting the J.I.C. is that of keeping senior levels of the Government informed about the imminence of hostilities on the best possible and most current basis.”¹⁵¹

Agreement

Ahead of the Bermuda conference, Macmillan wrote to Eisenhower to ensure that “Tripartite Alert Procedure” would be on the agenda for their meeting. In Bermuda, Macmillan and Eisenhower agreed that “[a]n effective machinery should be established for the rapid exchange of intelligence between the Governments of the United States, Canada and the United Kingdom on any sudden threat of Soviet aggression against the N.A.T.O. area. For this purpose a meeting between intelligence experts of the three Governments will be held in Washington immediately after the Bermuda Conference.”¹⁵²

After flying back from Bermuda, Allen Dulles, Dean, and Crean, along with advisers, met in Washington on March 28, 1957, for the first substantive discussion on how to achieve a tripartite agreement.¹⁵³ Dean began the meeting by describing the Paris meeting in 1954 and noting that current arrangements were not tripartite. While the British “had their letter from Allen Dulles to Mr. [Alan] Crick,” the UK’s liaison to the CIA, the procedure was “not really adequate for an arrangement of such importance.”¹⁵⁴ A tripartite arrangement, Dean went on, “should cover expeditious exchanges not only of information but also of assessments” (though he was careful to make clear he was in no way proposing tripartite agreed assessments). Dean presented the “draft formula,” which was the Canadian formula from the bilateral arrangement, suitably adapted and with slight modifications made by the British. This was to serve as the basis of an agreement between all three states.

Crean made clear that he supported the British position and noted that the bilateral arrangements confused things: it was unclear to one ally what the other two knew. He mentioned that any tripartite arrangement would augment but not replace the existing bilateral systems, which now included “direct line communications from C.I.A.” to the new Canadian Joint Indications Room. This was supplemented by liaison officers in Ottawa and Washington.

In a pointed interjection, Dulles said it might help if London had a “designated Indications Centre on the lines of those in Washington and Ottawa.” Dean “undertook to look into this.”¹⁵⁵ (Clearly, it was a good thing that Glazebrook had earlier pressed Ottawa to get to work on its own centre.) The meeting then shifted to discussion of how to physically pass intelligence and assessments among the three capitals. All agreed that “radio communications might be unreliable when most needed” and that intelligence should have its own channels, rather than having to compete with traffic along lines used by the services. Dean suggested that London might rent a channel in the new transatlantic cable, but that this was expensive; Crean suggested that Canada might pay half the cost. The details would have to wait.

Park Armstrong of the State Department pointed out that a “formal tripartite agreement” would be “breaking new ground” for the Department of State. He also worried about exclusion of other NATO allies. All the representatives agreed that informality was, in fact, helpful here: Crean and Dean both agreed that an exchange of letters between the Canadian and British ambassadors with the secretary of state would be all that was necessary.¹⁵⁶

By April 2, both the British official level and the IAC had agreed to the proposed arrangements. The State Department suggested that London and Ottawa proceed to send their formal letters.¹⁵⁷ Léger, in Ottawa, was “favourably surprised by the speed and receptiveness of the United States officials to the United Kingdom/Canadian proposal,” and assumed the speed was due to President Eisenhower’s direction to Dulles to “review the exchange of intelligence between the United States, United Kingdom, and Canada.”¹⁵⁸ The British foreign secretary, Selwyn Lloyd, approved the proposed agreements on April 8. Prime Minister Diefenbaker was shown the letter and had “no objection” by April 12, 1957.¹⁵⁹ The formal letters were given to given to Park Armstrong on April 16, 1957.¹⁶⁰

Less than a month later, on May 6, H. E. Furnas of the State Department gave advanced, if unofficial, notice that the US Government would approve the letters, with one major change of wording: “a likelihood of hostilities immediately threatening the NATO area”¹⁶¹ (the addition pertaining to immediacy). The British and Canadians also learned that the Americans would make their agreement contingent on an expectation the allies were operating their indications organizations on a twenty-four-hour basis (something that the Canadians assumed was “probably directed” at London where the British “do not maintain anything equivalent to our Joint Indications Room here or the National Indications Center in Washington”).¹⁶²

On May 8, Deputy Under Secretary Robert Murphy formally replied to both the British and Canadian ambassadors. On May 22 and May 31, the British wrote to the Canadians, and the Canadians replied to the British, respectively, “to complete the triangle.”¹⁶³ All parties agreed that the “arrangement should not repeat not be registered with the UN.”¹⁶⁴ The Tripartite Intelligence Alerts Agreement was in place.